Variant reduction method based on self-organizing mapping neural network

Abstract

The invention provides a variant reduction method based on a self-organizing mapping neural network, aiming at the problem that the number of variants in the field of variation testing is huge, so that the variation testing cost is too high. The method comprises the following steps: constructing a variation branch corresponding to a variant according to a weak variation conversion method, inserting the variation branch into an original program to obtain a new program, then executing the new program by using a test case set to obtain a variant killing matrix under a weak variation test, and taking the variant killing matrix as feature data of the variant; further acquiring the featuring variant position information and variation operator type information of variants to generate a variant feature matrix, and then adopting a self-organizing mapping neural network clustering technology for reducing redundant variants. According to the method, a large number of redundant variants can be reduced while the test effectiveness is hardly influenced, so that the variation test cost is reduced, and the software test efficiency is improved.

Description

technical field [0001] The invention belongs to the field of software testing, in particular to the field of variation testing, and relates to a variation reduction method based on a self-organizing mapping neural network. Background technique [0002] Software testing plays a vital role in ensuring software quality. Software testing exists in the entire life cycle of software development and has received extensive attention in the field of software reliability. Mutation testing is a basic software testing technique. The mutation test simulates the errors that may occur in the program. These errors are to make some small changes to a certain statement of the original program, such as replacing the relational operator "<" with "≤", injecting these errors into the original program to generate a new Program, the generated new program is called a variant, and the same test case is used to execute the original program and the variant. If the output of the program is different...

AI Technical Summary

Problems solved by technology

However, one of the main disadvantages of mutation testing is that a large number of variants are generated for a simple program, and most of the variants are redundant. A large number of variants increases the cost of mutation testing, so reduce redundant variants as much as possible. body, which can greatly improve test efficiency and reduce the cost of variation test

Images